I am working on household panel data in which there is a problem of attrition. i have calculated attrition weights by using inverse probability method. Now, when i run mixed command in Stata by using these weights, state shows the error as attrition weights were calculated against first wave. How to deal with this problem